The SmartVision™ Remote Diagnostics function enables operators to receive diagnostics information from their rolling stock fleets and visualize it in real-time. The control center, work planners and maintenance staff are continuously informed about the status of the fleet and potential technical issues. The staff can take actions instantly to fix the problems, limit their impacts on the operations as well as give the proper instructions to the driver even before he notices the problem.
Maintenance activities are expedited as the service personnel receives in advance information about issues which need actions. Thus, the maintenance scheduling and resource allocations are performed much faster, reducing service disruption and increasing both safety and operational efficiency.
SmartVision™ Condition Monitoring is used for detecting changes in the condition of such devices, for which Condition-Based Maintenance (CBM) shall be applied. The basic principle is to analyse and monitor changes in signals, which can be obtained from
- the existing train diagnostics system (e.g., temperature, pressure or voltage) or
- by adding new sensors to receive more detailed information about critical components (e.g. vibration sensors to monitor axle bearings or traction motors).
The technically and commercially optimal solution for each individual fleet type is defined together with our customers in order to best support them in their transition towards CBM.
SmartVision™ Condition Monitoring can be used to monitor the condition of the rolling stock, the track or both of them.
Examples of CIs include:
Adaptive Anomaly Detector
The condition monitoring and predictive analytics functionalities of SmartVision™ build upon Humaware’s Adaptive Anomaly Detector; a unique data-driven toolset that provides automatic, reliable and early alerts.
Its features include:
- More reliable and earlier detections than static thresholds can produce
- Detects gradual and sudden changes
- Adaptive thresholds remove the need to set up and maintain static thresholds
- Dramatic reduction in false alert rate across all asset types
- Detects changes earlier without increasing the false alert rate; a prerequisite for effective prognostics
- Result of 10+ years of research, development and field experience
SmartVision™ Predictive Analytics takes condition monitoring one step further by using advanced algorithms and analytics to predict likely remaining useful life of units. The SmartVision™ prognostics capability is provided by Humaware (an EKE Company).
The remaining useful lifetime estimates can be calculated for those components where the historical data collected from the vehicles indicate that the condition degrades against time or service cycles.
Predictive maintenance will enable optimisation of maintenance costs – when performing traditional preventive maintenance, servicing parts too early may unnecessarily shorten parts lifetime and increase costs. Conversely, if waiting too long, the component may experience failure and incur costly corrective maintenance and possible disruption in operations. Predictive maintenance addresses this by estimating remaining useful life using risk-based analysis techniques to optimise the time of maintenance.
Rolling Stock & Track Topology
SmartVision™ contains a hierarchical topology model of both the rolling stock and the rail network. The rolling stock topology stores information about the configuration and maintenance history of each vehicle and its subcomponents. The track topology, when correlated with SmartVision™ on-vehicle GPS information, enables the location of track related events to be pin-pointed.